m53297601
/
Paddle
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

Merge pull request #11874 from panyx0718/move_trainer

Browse Source 
Move trainer and utils api
revert-12383-port_py3_syntax
Xin Pan 7 years ago committed by GitHub
parent f7fd711e3f d70a38d8ec
commit 71b1c397d7
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
356 changed files with 543 additions and 533 deletions
Show Stats Download Patch File Download Diff File

2
README.md
Unescape View File

@ -18,6 +18,8 @@ learning to many products at Baidu.
Our vision is to enable deep learning for everyone via PaddlePaddle.
Please refer to our [release announcement](https://github.com/PaddlePaddle/Paddle/releases) to track the latest feature of PaddlePaddle.
### Lastest PaddlePaddle Version: [Fluid](https://github.com/PaddlePaddle/Paddle/tree/develop/paddle/fluid)
## Features
- **Flexibility**

2
doc/v2/design/cluster_train/large_model_dist_train.md
Unescape View File

@ -52,7 +52,7 @@ In `trainer_internal.cpp:L93 trainOneBatch`:
When doing actual network forward and backward, at the beginning of each batch, the trainer will try to download one row of data from pserver.
In `trainer/RemoteParameterUpdater.cpp`: `parameterUpdater_->getParametersRemote();`:
In `legacy/trainer/RemoteParameterUpdater.cpp`: `parameterUpdater_->getParametersRemote();`:
```c++
if (fullSize) {

40
doc/v2/design/mkl/mkldnn.md
Unescape View File

@ -18,20 +18,20 @@ Figure 1. PaddlePaddle on IA
具体的完成状态可以参见[这里](https://github.com/PaddlePaddle/Paddle/projects/21)。
## Contents
- [Overview](#overview)
- [Actions](#actions)
- [CMake](#cmake)
- [Matrix](#matrix)
- [Layers](#layers)
- [Activations](#activations)
- [Parameters](#parameters)
- [Gradients](#gradients)
- [Unit Tests](#unit-tests)
- [Python API](#python-api)
- [Benchmarking](#benchmarking)
- [Others](#others)
- [Design Concerns](#design-concerns)
- [Overview](#overview)
- [Actions](#actions)
- [CMake](#cmake)
- [Matrix](#matrix)
- [Layers](#layers)
- [Activations](#activations)
- [Parameters](#parameters)
- [Gradients](#gradients)
- [Unit Tests](#unit-tests)
- [Python API](#python-api)
- [Benchmarking](#benchmarking)
- [Others](#others)
- [Design Concerns](#design-concerns)
## Overview
@ -218,20 +218,20 @@ if use_mkldnn
我们总结出一些特别需要注意的点：
1. 使用**deviceId_**。为了尽可能少的在父类Layer中添加变量或者函数，
我们决定使用已有的`deviceId_`变量来区分layer的属性，定义`-2`为`MKLDNNLayer`特有的设备ID。
2. 重写父类Layer的**init**函数，修改`deviceId_`为`-2`，代表这个layer是用于跑在MKL-DNN的环境下。
我们决定使用已有的`deviceId_`变量来区分layer的属性，定义`-2`为`MKLDNNLayer`特有的设备ID。
2. 重写父类Layer的**init**函数，修改`deviceId_`为`-2`，代表这个layer是用于跑在MKL-DNN的环境下。
3. 创建`MKLDNNBase`，定义一些除了layer和memory相关的类和函数。
包括MKL-DNN会用到`MKLDNNStream`和`CPUEngine`，和未来可能还会用到`FPGAEngine`等。
包括MKL-DNN会用到`MKLDNNStream`和`CPUEngine`，和未来可能还会用到`FPGAEngine`等。
4. 如果MKL-DNN layer的后面接有cpu device，那么就会使`output_.value`与`extOutVal_`共享内存，
同时数据格式就是`NCHW`，这样下一个cpu device就能拿到正确的数据。
在有普通的CPU layer时， `extOutVal_`和`extOutGrad_`的格式始终是`NCHW`或者`NC`。
## References
1. [MKL small library](https://github.com/01org/mkl-dnn#linking-your-application)是[Intel MKL](https://software.intel.com/en-us/mkl)的一个子集。
主要包括了深度学习相关的数学原语与操作，一般由MKL-DNN在发布[新版本](https://github.com/01org/mkl-dnn/releases)时一起更新。
主要包括了深度学习相关的数学原语与操作，一般由MKL-DNN在发布[新版本](https://github.com/01org/mkl-dnn/releases)时一起更新。
2. [MKL-DNN System Requirements](https://github.com/01org/mkl-dnn#system-requirements)。
目前在PaddlePaddle中，仅会在支持AVX2指令集及以上的机器才使用MKL-DNN。
3. [原来的方案](https://github.com/PaddlePaddle/Paddle/pull/3096)会引入**nextLayer**的信息。
但是在PaddlePaddle中，无论是重构前的layer还是重构后的op，都不会想要知道next layer/op的信息。
但是在PaddlePaddle中，无论是重构前的layer还是重构后的op，都不会想要知道next layer/op的信息。
4. MKL-DNN的高性能格式与PaddlePaddle原有的`NCHW`不同(PaddlePaddle中的cuDNN部分使用的也是`NCHW`，所以不存在这个问题)。
所以需要引入一个转换方法，并且只需要在必要的时候转换这种格式，才能更好的发挥MKL-DNN的性能。
所以需要引入一个转换方法，并且只需要在必要的时候转换这种格式，才能更好的发挥MKL-DNN的性能。

2
doc/v2/dev/new_layer_en.rst
Unescape View File

@ -339,7 +339,7 @@ If you are creating a new file for the test, such as :code:`paddle/legacy/gserve
Implement Python Wrapper
========================
Implementing Python wrapper allows us to use the added layer in configuration files. All the Python wrappers are in file :code:`python/paddle/trainer/config_parser.py`. An example of the Python wrapper for fully connected layer is listed below. It has the following steps:
Implementing Python wrapper allows us to use the added layer in configuration files. All the Python wrappers are in file :code:`python/paddle/legacy/trainer/config_parser.py`. An example of the Python wrapper for fully connected layer is listed below. It has the following steps:
- Use :code:`@config_layer('fc')` at the decorator for all the Python wrapper class. :code:`fc` is the identifier of the layer.
- Implements :code:`__init__` constructor function.

4
paddle/CMakeLists.txt
Unescape View File

@ -1,7 +1,7 @@
if(NOT WITH_FLUID_ONLY)
add_subdirectory(legacy/cuda)
add_subdirectory(legacy/function)
add_subdirectory(utils)
add_subdirectory(legacy/utils)
add_subdirectory(legacy/math)
add_subdirectory(legacy/gserver)
add_subdirectory(legacy/parameter)
@ -10,7 +10,7 @@ if(NOT WITH_FLUID_ONLY)
add_subdirectory(legacy/capi)
else()
add_subdirectory(legacy/pserver)
add_subdirectory(trainer)
add_subdirectory(legacy/trainer)
add_subdirectory(scripts)
if(WITH_C_API)

2
paddle/fluid/operators/positive_negative_pair_op.h
Unescape View File

@ -14,7 +14,7 @@ limitations under the License. */
#include <vector>
#include "paddle/fluid/framework/eigen.h"
#include "paddle/fluid/framework/op_registry.h"
#include "paddle/utils/Logging.h"
#include "paddle/legacy/utils/Logging.h"
namespace paddle {
namespace operators {

2
paddle/fluid/platform/float16_test.cu
Unescape View File

@ -15,7 +15,7 @@ limitations under the License. */
#include "paddle/fluid/framework/lod_tensor.h"
#include "paddle/fluid/framework/tensor_util.h"
#include "paddle/utils/Logging.h"
#include "paddle/legacy/utils/Logging.h"
#define ARITHMETIC_KERNEL(op_type, sign) \
__global__ void op_type(const half* in1, const half* in2, half* out) { \

2
paddle/legacy/api/ConfigParser.cpp
Unescape View File

@ -14,7 +14,7 @@ limitations under the License. */
#include "PaddleAPI.h"
#include "PaddleAPIPrivate.h"
#include "paddle/trainer/Trainer.h"
#include "paddle/legacy/trainer/Trainer.h"
struct ParameterConfigPrivate {
paddle::ParameterPtr parameter;

2
paddle/legacy/api/Paddle.i
Unescape View File

@ -198,5 +198,5 @@ namespace std {
%ignore ParameterConfigPrivate;
%ignore OptimizationConfigPrivate;
%ignore ParameterTraverseCallbackPrivate;
%include "utils/GlobalConstants.h"
%include "legacy/utils/GlobalConstants.h"
%include "legacy/api/PaddleAPI.h"

4
paddle/legacy/api/PaddleAPI.h
Unescape View File

@ -20,8 +20,8 @@ limitations under the License. */
#include <string>
#include <vector>
#include "paddle/legacy/gserver/gradientmachines/GradientMachine.h"
#include "paddle/utils/Common.h"
#include "paddle/utils/GlobalConstants.h"
#include "paddle/legacy/utils/Common.h"
#include "paddle/legacy/utils/GlobalConstants.h"
/// Import PaddlePaddle's enumeration into global namespace.
using namespace paddle::enumeration_wrapper; // NOLINT

2
paddle/legacy/api/PaddleAPIPrivate.h
Unescape View File

@ -17,7 +17,7 @@ limitations under the License. */
#include "paddle/legacy/gserver/evaluators/Evaluator.h"
#include "paddle/legacy/gserver/gradientmachines/GradientMachine.h"
#include "paddle/legacy/parameter/ParameterUpdaterBase.h"
#include "paddle/trainer/TrainerConfigHelper.h"
#include "paddle/legacy/trainer/TrainerConfigHelper.h"
struct GradientMachinePrivate {
std::shared_ptr<paddle::GradientMachine> machine;

6
paddle/legacy/api/ParameterUpdater.cpp
Unescape View File

@ -16,10 +16,10 @@ limitations under the License. */
#include "PaddleAPIPrivate.h"
#ifndef PADDLE_WITHOUT_GOLANG
#include "paddle/trainer/NewRemoteParameterUpdater.h"
#include "paddle/legacy/trainer/NewRemoteParameterUpdater.h"
#endif
#include "paddle/trainer/RemoteParameterUpdater.h"
#include "paddle/trainer/ThreadParameterUpdater.h"
#include "paddle/legacy/trainer/RemoteParameterUpdater.h"
#include "paddle/legacy/trainer/ThreadParameterUpdater.h"
ParameterUpdater::ParameterUpdater() : m(new ParameterUpdaterPrivate()) {}

2
paddle/legacy/api/SequenceGenerator.cpp
Unescape View File

@ -19,7 +19,7 @@ limitations under the License. */
#include "PaddleAPI.h"
#include "paddle/legacy/gserver/gradientmachines/GradientMachine.h"
#include "paddle/legacy/parameter/Argument.h"
#include "paddle/utils/Flags.h"
#include "paddle/legacy/utils/Flags.h"
// used to represent partial sequence
struct Path {

8
paddle/legacy/api/Trainer.cpp
Unescape View File

@ -20,10 +20,10 @@ limitations under the License. */
#include <memory>
#include "paddle/legacy/gserver/gradientmachines/NeuralNetwork.h"
#include "paddle/trainer/ParamUtil.h"
#include "paddle/trainer/Trainer.h"
#include "paddle/trainer/TrainerInternal.h"
#include "paddle/utils/Flags.h"
#include "paddle/legacy/trainer/ParamUtil.h"
#include "paddle/legacy/trainer/Trainer.h"
#include "paddle/legacy/trainer/TrainerInternal.h"
#include "paddle/legacy/utils/Flags.h"
using paddle::real;

8
paddle/legacy/api/Util.cpp
Unescape View File

@ -15,10 +15,10 @@ limitations under the License. */
#include "PaddleAPI.h"
#include "paddle/legacy/parameter/Parameter.h"
#include "paddle/utils/Common.h"
#include "paddle/utils/Flags.h"
#include "paddle/utils/PythonUtil.h"
#include "paddle/utils/Util.h"
#include "paddle/legacy/utils/Common.h"
#include "paddle/legacy/utils/Flags.h"
#include "paddle/legacy/utils/PythonUtil.h"
#include "paddle/legacy/utils/Util.h"
#include <algorithm>
#include <iostream>

6
paddle/legacy/capi/Main.cpp
Unescape View File

@ -18,9 +18,9 @@ limitations under the License. */
#include <vector>
#include "capi_private.h"
#include "main.h"
#include "paddle/trainer/TrainerConfigHelper.h"
#include "paddle/utils/Excepts.h"
#include "paddle/utils/PythonUtil.h"
#include "paddle/legacy/trainer/TrainerConfigHelper.h"
#include "paddle/legacy/utils/Excepts.h"
#include "paddle/legacy/utils/PythonUtil.h"
static void initPaddle(int argc, char** argv) {
paddle::initMain(argc, argv);

2
paddle/legacy/capi/tests/test_Arguments.cpp
Unescape View File

@ -15,7 +15,7 @@ limitations under the License. */
#include <functional>
#include "capi.h"
#include "gtest/gtest.h"
#include "paddle/utils/ThreadLocal.h"
#include "paddle/legacy/utils/ThreadLocal.h"
static std::vector<paddle_real> randomBuffer(size_t bufSize) {
auto& eng = paddle::ThreadLocalRandomEngine::get();

4
paddle/legacy/capi/tests/test_GradientMachine.cpp
Unescape View File

@ -14,12 +14,12 @@ limitations under the License. */
#include <gtest/gtest.h>
#include <paddle/legacy/gserver/gradientmachines/GradientMachine.h>
#include <paddle/trainer/TrainerConfigHelper.h>
#include <paddle/legacy/trainer/TrainerConfigHelper.h>
#include <stdlib.h>
#include <string.h>
#include <type_traits>
#include "capi.h"
#include "paddle/utils/ThreadLocal.h"
#include "paddle/legacy/utils/ThreadLocal.h"
static std::vector<paddle_real> randomBuffer(size_t bufSize) {
auto& eng = paddle::ThreadLocalRandomEngine::get();

2
paddle/legacy/cuda/include/hl_base.h
Unescape View File

@ -208,7 +208,7 @@ typedef struct {
#include <cuda_runtime.h>
#include "paddle/legacy/cuda/include/hl_cuda.h"
#include "paddle/utils/Logging.h"
#include "paddle/legacy/utils/Logging.h"
extern __thread bool g_sync_flag;
extern __thread cudaStream_t default_stream;

2
paddle/legacy/cuda/include/hl_gpu_gru.cuh
Unescape View File

@ -18,7 +18,7 @@ limitations under the License. */
#ifdef __NVCC__
#include "paddle/utils/Logging.h"
#include "paddle/legacy/utils/Logging.h"
/*
* threads(framePerBlock, batchPerBlock)

2
paddle/legacy/cuda/include/hl_gpu_lstm.cuh
Unescape View File

@ -18,7 +18,7 @@ limitations under the License. */
#ifdef __NVCC__
#include "paddle/utils/Logging.h"
#include "paddle/legacy/utils/Logging.h"
#include "hl_device_functions.cuh"
/*

2
paddle/legacy/cuda/include/hl_gpu_matrix_kernel.cuh
Unescape View File

@ -18,7 +18,7 @@ limitations under the License. */
#define HL_GPU_MATRIX_KERNEL_CUH_
#include <algorithm>
#include "paddle/utils/Logging.h"
#include "paddle/legacy/utils/Logging.h"
#include "hl_base.h"
#ifdef __NVCC__

2
paddle/legacy/cuda/src/hl_cuda_aggregate.cu
Unescape View File

@ -18,7 +18,7 @@ limitations under the License. */
#include "hl_cuda.ph"
#include "hl_matrix_base.cuh"
#include "hl_thread.ph"
#include "paddle/utils/Logging.h"
#include "paddle/legacy/utils/Logging.h"
/**
* @brief matrix row operator.

4
paddle/legacy/cuda/src/hl_cuda_cublas.cc
Unescape View File

@ -16,8 +16,8 @@ limitations under the License. */
#include <sys/time.h>
#include "hl_cuda.h"
#include "hl_thread.ph"
#include "paddle/utils/DynamicLoader.h"
#include "paddle/utils/Logging.h"
#include "paddle/legacy/utils/DynamicLoader.h"
#include "paddle/legacy/utils/Logging.h"
namespace dynload {

4
paddle/legacy/cuda/src/hl_cuda_cudnn.cc
Unescape View File

@ -17,8 +17,8 @@ limitations under the License. */
#include <gflags/gflags.h>
#include "hl_cuda_cudnn.ph"
#include "hl_thread.ph"
#include "paddle/utils/DynamicLoader.h"
#include "paddle/utils/Logging.h"
#include "paddle/legacy/utils/DynamicLoader.h"
#include "paddle/legacy/utils/Logging.h"
DEFINE_int32(cudnn_conv_workspace_limit_in_mb,
4096,

Some files were not shown because too many files have changed in this diff Show More

Write Preview
Loading…
Cancel
Save